ANNALES Annales Zoologici (1994) 45: 15-26 ZOOLOGICI

Emancipation of Slaves in sanguinea Latr. Colonies (, Formicidae)

Wojciech CZECHOWSKI

Museum and Institute of Zoology, Polish Academy of Sciences, Warsaw, Poland

A b stra c t. Formica sanguinea Latr. occasionally takes wood as its slaves. Eleven nests of the former were experimentally provided with various amounts (10,000-250,000) of pupae of F. polyctena Foerst. or F. rufa L., and thus, mixed colonies with different proportions of slaves (29-96%) were obtained. The slaves destroyed the F. sanguinea colonies by elim inating their brood and queen(s) (with no aggressiveness towards adult workers). The following year, even though F. sanguinea workers were still present, the slaves adopted young queens of their own species. Before their complete emancipation, the slaves fought on the side of the slavers during conflicts with conspecific ants, even against their former nestmates. Emancipation restored mutual comity between the ex-slaves and the unenslavedF. polyctena. At the same time, the ex-slavers lost their aggressiveness towards alien F. polyclena (but not vice versa). The mixed colonies were raided by foreign F. sanguinea. The studies were carried out in the Gorce Mts (the Western Carpathians) from 1987 to 1991.

Key words: Formica sanguinea, wood ants, social , dulosis, mixed colonies, queen adoption, nest takeover, aggressiveness, colony odour

INTRODUCTION thorpe 1915, Wheeler 1926, Forel 1928, Dobrzański 1961, Kutter 1968, Czechowski 1975a) or by another Formica (Raptiformica) sanguinea Latr. is a species. Such atypical slaves, F. pratensis Retz. common northern-Palaearctic slave-maker. Ants of and F. polyctena Foerst., were found during recent the subgenus Seruiformica For. are its typical studies on the myrmecofauna of Polish pine forests slaves. Occasionally, however, particularly in Cen­ (Tucholskie Forests, Biała Forest) in two out of tral Europe, this function is fulfilled by wood ants about 80 (2.5%) recorded nests of F. sanguinea (the subgenus Formica s. str.) (Stitz 1914, Donis- (Czechowski et al. 1995). Their shares in the mixed colonies were over a dozen percent (Czechowski, Address for correspondence: Wojciech Czechowski, Muzeum i Pisarski, unpubl. data). Podkówka (pers. comm.) Instytut Zoologii PAN, Wilcza 64, 00-679 Warszawa, Poland http://rcin.org.pl 16 • W. Czechowski

Table 1. Range, course and results of the experiments on slave emancipation in Formica sanguinea colonies

Original numbers of Slart of the Pupae introduced Proportion of slaves Colony F. sanguinea experim ent numbers species in autumn (%) SP-I 8,000 1987 230,000 F. polyctena 96 SP-II 50,000 1988 250,000 F. polyctena 80 SP-III 20,000 1989 25,000 F. polyctena 52 SP-IV 1989 110,000 F. polyctena 50,000 66 SP-V 1989 15,000 F. polyctena SP-VI 4,000 1989 30,000 F. polyctena 85 SP-VII 10,000 1989 50,000 F. polyctena 88 SP-VIII 40,000 1989 20,000 F. polyctena 29 SP-IX 2,000 1990 10,000 F. polyctena 82 SP-X 5,000 1990 50,000 F. polyctena 90 SR-I 4,000 1989 15,000 F. rufa 77 SF-I ? 1989 20,000 F. fusca 36 observed numerous natural colonies of F. sanguinea tendency to take unlimited numbers of alien pupae mixed with F. pratensis at Prószków (Silesia), as supplied (Czechowski 1989, 1990a). [Marikovsky well as with ants of the subgenus Coptoformica (1963), however, reported that the pillaging activity Mull, at Złocieniec (Pomerania). Galle(pers. comm.) of F. sanguinea weakened when there was a lot of saw even Camponotus fallax (Nyl.) asF. sanguinea slaves present in their nests]. This very greed of the slaves in Hungary. No phenomena like these seem slavers turns against them when wood ants, for to occur in other regions [Finland: Rosengren et al. instance F. pratensis (Forel 1874, 1928) or F. poly- (1986); Siberia: Bugrova, pers. comm.]. There may ctena (Czechowski 1990a), are their slaves. These, be some zoogeographic regularity in this case be­ when occurring in great numbers, destroy an entire cause, for instance, in Great Britain (this country F. sanguinea colony by eliminating its queen(s) and being the limit of the range of the species), F. san­ brood. Later, they adopt young conspecific queens guinea takes adventage of slaves (even typical ones) and the colony of F. sanguinea is gradually trans­ to a lesser degree than in Central Europe (Darwin formed into an ordinary colony of the former victim 1859, Sudd, Franks 1987). A similar situation oc­ (Czechowski 1990a). A different scenario is also curs in Central Asia (Tien Shan Mts; Marikovsky possible, as the one described by Marikovsky (1963) 1963) - an area of isolated populations of the from Western Siberia, with F. rufa L. as a slave species. species: swarms of slaves and slavers were sepa­ The abundance of slaves in F. sanguinea nests rated and this led to the origin of something like a depends on colony size and on the availability of bispecific poly(oligo-?)calic colony. potential prey; F. sanguinea appears to show a The present paper is a continuation of earlier studies (Czechowski 1989, 1990a,b). The objective was to investigate the phenomenon of slave eman­ 1 The term “slaves” used for these individuals may seem inappropriate because of their species status not typical of cipation1 in a quantitative aspect, into the concomi­ ordinary slaves, their enormous relative abundance, which usually highly exceeds that of theF. sanguinea colonies, and the tant changes in the ant behaviour and to possibly artificial manner in which these ants entered a given colony (and explain the discrepancies between the observations also their fatal impact). The term “emancipation” used to define the phenomenon described is controversial, too. Hovewer, it is of Forel and the author on the one hand, and those difficult to find justification for abandoning the traditional term of Marikovsky on the other. “slaves”. Wood ants can be slaves ofF. sanguinea under natural conditions, the typical (and natural) slaves ( Serviformica) occasionally have an abundance highly exceeding that of the colony of their parasite, and the very difference in the manner in which foreign pupae get into a nest of F. sanguinea (natural kidnapping ana experimental introduction) cannot be a sufficient reason for treating both groups as different qualitative categories. The term “emancipation” is therefore simply the logical consequence of keeping the term “slaves”

http://rcin.org.pl Emancipation of Slaves in Formica sanguinea Latr. Colonies • 17

Proportion ofF. sanguinea the next year (%) Adoption Colony status the following years (till 1991) spring sum mer autum n of new queens 0 0 0 + pure F. polyctena <1 0 0 + pure F. polyctena 3 0 2 + mixed F. polyctena + F. sanguinea

5 3 0 + pure F. polyctena

8 3 ? + pure F. polyctena 9 colony included to Colony SP-III 23 13 ? ? stroyed by foreign F. sanguinea 0 ? ? + pure F. polyctena <1 ? ? + pure F. polyctena <1 0 0 + mixed F. rufa + F. polyctena 80 ? ? - normal F. sanquinea + slaves

STUDY AREA, MATERIAL, METHODS initial abundance of a f. sanguinea colony (Table 1) [assuming that there had been some insignificant The studies were carried out in the Jaszcze losses of pupae; see: Czechowski •(1989)]. 2 stream valley (locality of Ochotnica Górna) in the F. polyctena pupae introduced into a given colony Gorce Mts (the Western Carpathians, southern Po­ of F. sanguinea came (depending on their number) land) from 1987 to 1991. The experiments were from one or more nests (mainly from the same conducted on 11 colonies of various sizes (2,000- polycalic colony). They were provided (also depend­ 50,000 individuals) of F. sanguinea, which were ing on their number) in one or several batches, and provided with different numbers (10,000-250,000) this ensured that each time there was enough room of pupae of F. polyctena workers (10 cases; Colonies for them in the nest (see: Czechowski 1989). F. rufa SP-I-X) or F. rufa ones (Colony SR-I). [Some find­ pupae came from four monocalic nests and were ings (data from Colonies SP-I and SP-II) were pub­ provided in four (mixed) batches. The F. sanguinea lished earlier as a preliminary report (Czechowski colonies were supplied with pupae from the end of 1990a)]. Pupae from mounds of wood ants were June till August, but mainly in July. The percentage collected by means of wax honeycomb frames (Wi­ of slaves was determined after their entire emer­ śniewski 1973); their number was estimated by the gence in autumn (the end of September), then in volumetric method (1 litre = 30,000 pupae). The spring (April-May), summer (July) and autumn of number of pupae introduced into a nest and the the following year. In order to do this, random percentage of slaves that emerged in a mixed society samples of ants were collected from the surface of a provided a basis for determining an approximate nest after disturbing the colony a little. Regular observations covered the experimental colonies and 2 In colonies of amazon ants, eg Polyergus lucidus Mayr, most of the nearby normal colonies of F. sanguinea the pupae captured may be devoured (Cool-Kwait, Topoff 1984). However, in amazon ants, unlike in F. sanguinea, foraging is throughout the season. exclusively the duty of slaves and therefore it is no wonder that Apart from this, as a control, one colony of F. food maybe in short supply. Furthermore, pupae captured by slave-keeping ants themselves come to the nest gradually and in sanguinea (SF-I) was provided with F. fusca L. batches smaller than those provided experimentally. Thu s, u nder pupae. These were collected, by means of a vacuum natural conditions far fewer of them have any chance of undergoing complete metamorphosis. Nevertheless, the values cleaner, from many nests from mid-July to mid-Sep­ of the initial colony size ofF. sanguinea obtained on the basis of the percentages of slavers and slaves in ant samples collected tember. from the surface of (slightly disturbed) nests must be treated with One of the colonies (SP-V) appeared to be a great caution - simply as indicatory values. Such data could be entirely reliable only on the assumption (out of the question here) temporary branch of another one (SP-IV); all the that the percentage of the emerged slaves was identical in each of the nests, and that the surface samples represented the pupae provided there were carried by F. sanguinea proportion of species real for the whole colony. to its main nest. Colonies SP-I, -II, -VIII, and SR-I

http://rcin.org.pl 18 • W. Czechowski were situated far from one another. However, Col­ and devouring their offspring. In colonies with a onies SP-III and SP-VII, and also SP-IV, -VI, -IX, -X relatively small percentage of slaves (SP-VIII, -IV, and SF-I were very near. The latter were a part of -III)F. sanguinea survived a little longer (Table 1), a larger F. sanguined complex. Yet all the colonies and this may have been due to incomplete annihila­ formed separate (monocalic) societies. They did not tion of the new generation of its workers. It must be keep in contact and alien workers (confronted ex­ underlined here that there were recorded absolutely perimentally) displayed something of an aggressive­ no signs of hostility of slaves (and later ex-slaves) ness (dragging) towards one another. towards adult F. sanguinea workers throughout o As a result there were 10 mixed colonies in which their life together. The latter died out naturally. wood ants were the slaves of F. sanguinea (9 nests The two presumed mechanisms of emancipation with F. polyctena and one with F. rufa). The initial (elimination of the queens and brood of the slavers) proportions of the slaves were comprised within a were (at least in great part) confirmed when two large range: from 29% (SP-VIII) to 96% (SP-I) (Table experimental nests, SP-IX and SP-X, had been dug 1). Almost all the F. sanguinea colonies studied out. This was done in autumn (7 September, 1990), contained “natural” slaves of the genus Seruiformi- more than two months after the introduction of F. ca: F. cunicularia Latr. and (or) F. fusca. Their polyctena pupae (18 June into SP-IX; from 18 June insignificant proportions of these (only a few per to 4 July into SP-X). No queens were found in each cent) had no bearing on the calculations or the of the nests (of course, this can not be regarded as course of the experiments. an uncontrovertible evidence that the colonies were really queenless, especially if they were mono- gynous). Only fourF. sanguinea pupae were found RESULTS in SP-IX and there was no brood at all in SP-X. Before the introduction of slaves, both colonies had Slave em ancipation had their own offspring and in September all the Emancipation of slaves occurred in all or almost nearby untouched nests of F. sanguinea were full of all the SP and SR colonies (only in the case of Colony their own pupae. SP-VIII the fact of emancipation was not recorded The case of Colony SP-III is puzzling. In 1989, mostcertainly). Duringthe springofthe secondyear the initial share of slaves there was 52%. In the of the experiments the swarms of the social parasite following spring (1990) F. polyctena already con­ gradually disappeared and F. sanguinea queens stituted 97%, and in autumn 98%. There emerged a were replacedbyyoungqueens of wood ants adopted new generation of workers - a sign that a queen after their nuptial flights by orphaned swarms of (queens) had been adopted. Meanwhile, in spring (former) slaves. The adoption of queens was dis­ 1991, the share of F. sanguinea workers increased covered either directly (SP-I, SR-I) or indirectly, on to 14%. The difference in the ratios of both species the basis of the emergence of brood (a new worker in this sample (n = 118) and in the previous one generation) of wood ants in (former) nests of F. (n = 60) is statistically significant (rf-test; P = 0.01). sanguinea. Colonies SP adopted conspecific queens Therefore, readoption of a F. sanguinea queen can­ and developed (or, at least, started to develope) into not be ruled out. Maybe, she had survived in the congeneric societies of F. polyctena. The oldest (and vicinity of the nest, together with a group of her own the biggest) of them, SP-I and SP-II, even began workers which still were in some contact with the producing their own alate queens (starting from rest of the mixed colony. Since something of the kind 1990). Colony SR-I simultaneously adopted F. rufa of bispecific oligocalic colonies of F. sanguinea + F. queens (conspecific to the ex-slaves) and F. poly­ rufa (Marikovsky 1963) and of ctena ones, thus becoming a potentially permanent (Latr.)-ł-F. cinerea Mayr (Czechowski 1975b) which bispecific society (Czechowski 1991b). function on the principle of oligogyny are possible, As concluded earlier (Czechowski 1990a), the it follows that a monocalic oligogynous colony of this emancipation of slaves was achieved through elimi­ nation in some way (killing?) of F. sanguinea queens 3 In species of the genus Camponotus Mayr the adopted alien brood kill the host workers (Carlin 1988).

http://rcin.org.pl Emancipation of Slaves in Formica sanguined Latr. Colonies • 19

type is also conceivable (although it seems rather colony. Towards the end of the whole operation F. unlikely). Forel (1874, 1928), too, recorded cases of sanguinea revealed their conservative attitude. mixed colonies of F. sanguinea + F. pratensis surviv- They showed a tendency to remain in the deserted ingfor several years without any intake of new slave nest, to return to it or they even tried to change the pupae. He ascribed this to a very long life span of direction of the move. A spectacular attem pt at workers observed under laboratory conditions. forcing the swarm (still a mixed one) to return to its However, potential longevity and real longevity former nest was undertaken by F. sanguinea from under field conditions are two different things (Do­ Colony SP-VI after the emancipation of the slaves brzańska, Dobrzański, pers. comm.). F. sanguinea (during the following year, after the introduction of workers in the Gorce Mts lived for 20-22 months at pupae and removal to a new nest). Although they most (this included two winters). were in minority (8% of the colony), the ex-slavers No emancipation of slaves was recorded in Nest began, in June, a desperate transportation of the SF-I (with F. fusca). This experiment, regardless of ex-slaves to their old nest (2.5 m away). As a result its result, must be considered a failure. The F. of this, both nests were inhabited for some time sanguinea swarm consisted of no more than 5,000 until, in July, F. polyctena undertook a firm counter­ workers and the naturally enslavedF. fusca individ­ action that sealed the fate of the old nest, which was uals constituted 5% of the colony. In spite of the fact abandoned altogether. that 20,000 pupae were provided, the share of F. Nest conservatism in F. sanguinea is illustrated fusca increased only to 36%. Thus the losses must perfectly by the case of Colony SP-X. It had changed have been very high, and this was due to several places three times after the emergence of slaves and causes. It was difficult to collect such a great num­ before it was dug out. On the day it was dug out the ber of F. fusca pupae and therefore they were pro­ colony inhabited four nests, though to various de­ vided over a long period (nearly two months) in grees: the original nest of F. sanguinea (A) and the many (19) small (100-3,000 individuals) batches. successively changed new nests (B, C, D). The dis­ Frequently, the pupae were young and required a tances between them were 6.5, 3.0 and 3.4 m, re­ long time to emerge. About half of the total number spectively. All four were dug out and samples of ants consisted of nude pupae (without cocoons), which were collected. Nest A was almost completely were easily damaged during handling. Thus.F. san­ deserted - only 23 individuals were found there: 1 guinea had a possibility to eat more of these than of F. polyctena (4%) and 22 F. sanguinea (96%). Nest pupae of wood ants that were practically at the same B, too, was almost entirely empty - 25 individuals age, advanced in their development, always cocooned (nearly all) were caught: 8 F. polyctena (32%) and and provided in “shock” batches. 17 F. sanguinea (68%). Nests C and D were fully inhabited and joined by a route of intensive two-way Nest relocations traffic. In the samples taken from both nests (n = 59 In summer, soon after the beginning of the mass and 87) the ratio of F. polyctena and F. sanguinea emergence of slaves, the mixed colonies changed was identical - 91% : 9% (Fig. 1). The difference in their place of nesting or at least tried to do so. Only Colonies SP-IV and SR-I had given up their at­ tempts (building initiatory nests) to find new nest­ ing place and remained in their old ones (considerably enlarged and rebuilt). Colony SP-II moved the farthest (25 m) (Czechowski 1990a). The initiative to move or, more precisely speak­ ing (considering their later attitude), to found new filial nests, generally came from F. sanguinea work­ ers. Only later (a removal usually lasted for some Fig. 1. Nest conservatism of F. sanguinea and changes in the weeks) did great numbers of older slaves join the proportion of the slavers (white) to the slaves (black) in the operation. This decided the removal of the entire successively settled nests of the mixed Colony SP-X.

http://rcin.org.pl 20 • W. Czechowski the proportions is manifest. “Natural” slaves (F. which could be ascribed beyond doubt to F. sangui­ fusca), constituting 2% of the mixed colony, occurred nea (and larvae of this species should have been only in Nests C and D. present at that time). It is, therefore, evident that The situation was similar during removals of the brood of the slave-keepers were destroyed other colonies, eg SP-III (this, exceptionally, was (eaten) very early - at the beginning of a mass inspired by the slaves). After nearly a month-long emergence of slaves and before establishing new removal the old nest contained 19 ants: 5 F. poly- ways of foraging. ctena (26%) and 14 F. sanguined (74%). In the new A certain (small) number of F. sanguinea work­ nest this ratio was 52% : 48% (n = 81). The difference ers usually wintered in their deserted old nest. is statistically significant (d-test; P < 0.05). In this These joined the rest of the society in spring (eg case, too, all “natural” slaves (F. cunicularia), con­ SP-VI) or, if they had lost contact, died there in stituting 10% of the mixed colony, moved to the new isolation (eg SP-II; Czechowski 1990a). nest. In one respect the slavers managed to impose During an attempted but unrealized removal of their nesting habit on the slaves. In all the cases Colony SP-VII the ratio of F. polyctena to F. sangui­ whenF. sanguinea had a separate winter nest (SP-I, ned was 73% : 27% in = 41) in the initiatory nest -II, -IV, -VIII) the mixed colony, or at least a part of whereas it was 88% : 12% (n = 108) in the main one. it (SP-II; Czechowski 1990a) just wintered there The difference is significant (P < 0.05). even when their new nest (SP-I, -II) or that greatly In the light of these facts it can be concluded that enlarged rebuilt one (SP-IV) showed all the signs of F. sanguined, while undertaking the removals, did all-the-year-round nest of wood ants.5 Only in the not intend to change the place of nesting but wanted case of Colony SP-VIII still dominated by F. sangui­ to found filial nests, probably foraging stations.4 As nea (Table 1), did this seem justified. But even the soon as the colony size was suddenly and considera­ four-per-cent minority in SP-I forced an autumn bly increased, the ants undoubtedly began to experi­ removal of the entire colony to their winter nest ence a shortage of food. The nests of F. sanguined which had been left unused throughout the season. were situated in more or less open places where What is more, this nest had to be rebuilt very quick­ their carbohydrate food (honeydew) came from ly because its mound had been pulled down and the small and dispersed colonies of aphids on herba­ material used for the new nest (Czechowski 1989). ceous plants and on small spruces; this was not Nest conservatism ofF. sanguinea, recorded dur­ sufficient in the new situation. The slavers (because ing the experiments, may have been the cause of the their slaves were still too young to forage) were faced phenomena described by Marikovsky (1963) - the with the necessity to provide the greatly increased separation of swarms of the parasite and its victim. food requirements of the societies. While expanding As a result of such an interpretation the difference the range of foraging they came across rich sources in the gist of the phenomena recorded in the Gorce ofhoneydew in the canopies of old spruces. As a rule, Mts and in Western Siberia becomes invalid, the the new nests (stations?) were founded on the edge latter phenomenon being reduced to special instan­ of a forest or under bigger trees. These new localities ces of the former. If F. sanguinea queens had been were also better than the previous ones as nesting eliminated during the time when both species were places for wood ants (shade, abundance of suitable living together, their readoption could have been building material, a rich source of food close by). No provided a possible difference (overlooked, maybe, wonder they were preferred by the slaves who were by Marikovsky). gaining quantitative advantage. During all the observed removals adult workers and pupae were transported. There were no instan­ ces of carrying larvae (and queens), that is offspring,

5 Wood ants colonized artificially during late summer (or even in autumn; Czechowski, unpubl. data) in non-optimum habitats 4 It cannot be ruled out, however, that these were emancipation are able to prepare nests that ensure sate overwintering attem pts made by... F. sanguinea. (Pisarski, Czechowski 1990).

http://rcin.org.pl Emancipation of Slaves in Formica sanguinea Latr. Colonies • 21

Relations with other colonies The relations of the mixed experimental colonies Among all the experimental nests, only SP-III with foreign untouched F. sanguinea colonies were and SP-VII were situated close enough (9 m) to completely different. The latter attacked the former make direct contact of the foreign colonies possible. and treated them as sources of slaves for them­ Each of them occupied a heap of stones that had selves. Such attacks were directed at Colonies SP-VI been formed on a grassy slope. The heaps were old, (twice) and SP-VIII. The first attack on SP-VI took overgrown with thermophilous vegetation (dog- place one month after the colony had been provided rose, raspberry and blackberry bushes, wild straw­ with pupae, at the beginning of mass emergence of berries, and thyme) - typical habitats of F. F. polyctena. The course of the conflict was very sanguinea. The colonies had made no contacts be­ interesting and the whole affair was a failure and fore F. polyctena were introduced. Individuals ex­ the aggressor withdrew (see: Czechowski 1990b). perimentally relocated for control reasons were recognized as foreign (they made the local ones interested and agitated) but provoked no distinct aggressiveness. During the first year of the experiment, in mid- September of 1989, Colony SP-VII built a small initiatory nest in a stretch of grass dividing the sites

of the two nests. In this way the ants came closer (6 1 m m) to an old, still partly inhabited nest of Colony SP-III. This colony immediately reacted to a change Fig. 2. Aggression of F. sanguinea (N-II) towards the mixed in the status quo. A loose column of excited workers Colony SP-VI: A -old nest of SP-VI, now occupied by Colony N-II; started out towards the new nest, stopped 0.5 m B - new nest of SP-VI; C - old temporary nest of Colony N-III of before their target and formed a restless concentra­ F. sanguinea (see: Czechowski 1990b), now a shelter for a part of tion - as during an ordinary raid of F. sanguinea the ants from SP-VI; 1 - raid of F. sanguinea: direction of the attack, manoeuvre and chase; 2 - direction of the escape of Colony (Czechowski 1975a, 1977). F. sanguinea were the SP-VI. main participants of this operation although Colony SP-III was (slightly) dominated by F. polyctena Colony SP-VI was attacked for the second time then. No attack followed. The operation looked like in the following year (1990), after slave emancipa­ a warning given to a (known) neighbour who was tion. As had happened before, the alien F. sanguinea trespassing against the established norms of coex­ (N-II) first settled (at the end of July) in the near istence. On the following day the initiatory nest of vicinity - this time occupying a deserted old nest of Colony SP-VII was completely deserted. Colony SP-VI (Fig. 2).6 The aggresor attacked on 3 On the one hand, this incident pointed out to August, in the morning, and outflanked Nest SP-VI some connection (genetic?) between both colonies of on both sides. The attacked ants defended their F. sanguinea (which refrained from open aggres­ colony at the base and then on the mound surface. sion) but on the other hand, it stressed their current In the course of this, their evacuation route was separateness. The latter acquired special import­ quickly revealed and it led directly to a place 24 m ance in the light of a later event. Now, both these away - far beyond the regular range of the colony. colonies - of already emancipated slaves but still At first, pupae were transported only sporadically. with some F. sanguinea (Table 1) - fused at the end Then, at noon, a big group of F. polyctena (together of June of the following year. The fusion automati­ with the remainders of F. sanguinea) suddenly re­ cally included the remainders of both slave-keeper turned from the place of evacuation to the attacked swarms. The union was initiated by F. polyctena nest and initiated a mass flight with their pupae. from SP-VII who had first made contact with their Among those transported this time there were ob­ conspecifics from SP-III and then moved to their (new) nest situated 12 m away from Nest SP-VII. 6 Colonics of F. sanguinea move from one place to another in search of prey (Rosengren et al. 1986).

http://rcin.org.pl 2 2 • W. Czechowski served some big pupae which turned out to be pupae there was no trace of the experimental colony. In of F. sanguined malesd), which means the offspring autumn, however, there were recorded no F. poly­ of orphaned workers. (And they were being rescued ctena workers (emerged from allegedly captured byF. polyctena workers!). [SomeF. sanguinea work­ pupae) in the nest of the aggressor. But the foreign ers may have well developed ovaries (Móglich, Hóll- colony of F. sanguinea was very large whereas dobler 1974), however male production by workers pupae in SP-VIII were very few (species affiliation was not reported earlier in this species]. A moment ofthe latterhad not been checked). Itis also possible later the aggressors completely encircled the nest that the attacked ants managed to escape without and intercepted the rest of the evacuation column. great losses and they later built a new nest some­ The ants that had managed to break through were where else. chased for over 5 m and their pupae were taken An unusual form of kidnapping (eudulosis) was away. The whole of Nest SP-VI was captured. Col­ employed by F. sanguinea towards Colony SR-I at ony N-II managed to seize two-three thousand the turn of August and September 1990. That colony pupae; most had been taken away in time. After the was after the emancipation of a slave swarm of F. o combat from an area of 800 cm at the mound base rufa, already a mixed one - F. rufa + F. polyctena of Nest SP-VI dead ants were collected: 75 (66%) F. (see: Czechowski 1991b). Its age structure was not polyctena and 39 (34%) F. sanguinea. The total typical either - the old F. rufa workers (ex-slaves), number of casualties was not much higher. very few all along, were already dying out and the In its new place Colony SP-VI made a bivouac in new, very numerous mixed generation (the first a tuft of grass and the ants stayed there passively offspring of young queens) was only beginning to for three rainy days. When the weather improved emerge. Thus the majority of the colony consisted of they immediately begun searching for their nest- very young and weak workers which were forced, too mates, dispersed during the conflict around, their early, to undertake tasks outside the nest. Some F. (old) nest. A large group (with pupae) had gathered sanguinea workers (from an unidentified colony) under a stone 3 m away from the nest (in a deserted wandered singly round the nest and captured these temporary shelter of the previous year’s aggressor; callow ants (without doing them any harm). How­ Czechowski 1990b). The rescue operation was very ever, it is possible that, in the autumn, Colony SR-I competent. There was no doubt that it had been was openly attacked because by the following spring undertaken with just this intention. The ants that its nest was deserted. But the colony had tried to were involved in it picked their way among numer­ change its nesting place earlier because of the tire­ ous foreign F. sanguinea individuals penetrating some presence of a strong Lasius niger (L.) colony, the area - but these were ignored by the searchers manifesting an aggressive attitude. interested only in finding their lost nestmates. The relations between F. polyctena ants from the Colony SP-VI (additionally supplied with F. poly­ experimental and natural colonies depended on the ctena pupae) built a new nest, far from the old one current status of the former. They were definitely which had been taken over by the aggressors for hostile during the first year of the experiment (Cze­ good. Colony N-II retained its species status; the chowski 1989, 1990a) but normalized after slave number of F. polyctena pupae captured must have emancipation. For instance, between the initially been insufficient for emancipation of the slaves. An (1987) hostile Colonies K-III (natural) and SP-I, inspection carried out in September 1991 revealed during the second year (1988) there was formed a no presence of F. polyctena in Colony N-II. “no ants zone” (Czechowski 1990a); later on (1989- Colony SP-VIII was attacked by F. sanguinea 1990) a border impassable for both parties; then nesting about 20 m away. This occurred at the end (1991) the colonies formed a federation-SP-I moved of July 1990 when F. sanguinea workers were still into the territory of the stronger partner but kept quite numerous in the colony (Table 1). The raid was its individual character; finally (1992), Colony SP-I not recorded but its results were evident: Nest SP- moved into one of the two, at that time existing, VIII was occupied by alien F. sanguinea (and func­ nests of K-III, probably because its own nesting tioned as its branch till the end of the season) and place had become too overshadowed. This “re-

http://rcin.org.pl Emancipation of Slaves in Formica sanguinea Latr. Colonies • 23 straint” in developing comity should be associated concerning the origin, sources and biochemical na­ not so much with the origin of the ex-slave colony ture of these signals (Hólldobler, Michener 1980, but more with the time of its appearance within the Sudd, Franks 1987, Hólldobler, Wilson 1990). The range of Colony K-III. Artificially founded colonies most recent discoveries point out to surface phero- of F. polyctena, even of different origin, show a mones, particularly to cuticular hydrocarbons strong tendency to merge when they are colonized (CHC), as the probable olfactory discriminators (or during the same year. However, already stabilized the so-called colony labels) in various groups of colonies (for instance, established the previous sea­ social , including ants (Bonavita-Cougour- son) tend to preserve their separateness in relations dan, Clement 1986, Bonavita-Cougourdan et al. with younger ones, even of the same origin, and 1986, 1988, Yamaoka et al. 1988, 1990, 1991). The their initial contacts are more or less aggressive composition of the CHC mixture is species-specific (Pisarski, Czechowski 1990, Czechowski, unpubl. and genetically controlled. However, as any other data). The case discussed here, when Colony SP-I character, this is also subject to individual vari­ (already independent) appeared near K-III (a resi­ ability. This variability for a given colony falls with­ dent of the area), may be considered a special case in the range of the species variability and is the of artificial colonization (which, in fact, it was; Cze­ resultant of the odour variability of all the members chowski 1991a). of a given society. Individual odours are stand­ As mentioned above, new “artificial” colonies ardized (mixed) in the course of intracolonial inter­ merge without any problems during the first year of specific social contacts (Yamaoka 1990, 1991). The their existence; just like worker groups under role of the queen is still controversial: is she the stress. When Colony SP-VI, weakened after its con­ main source of the identifier of her colony (Carlin, flict with alien F. sanguinea, was staying in its Hólldobler 1983) or is her impact in this respect bivouac the ants were provided with F. polyctena reduced to her role as the centre of social life in pupae taken from the same polycalic colony from which the odour is standardized (Yamaoka 1990)? which pupae forF. sanguinea had been collected a The ability to recognize nestmates (and the atti­ year earlier. There were about 10,000 pupae togeth­ tude towards non-nestmates) is decided by the early er with several hundred workers and some nest experience of workers imprinting the odour stimu­ material. Not the slightest conflict between the ex­ lus from the social environment of the callow ants slaves and the new F. polyctena ants was recorded. (Jaisson 1975, 1980, Jaisson, Fresneau 1978, Le The strewn pupae were carried into the bivouac by Moli, Passetti 1977,1978, Le Moli, Mori 1982,1984). all of them side by side. But SP-VI still contained All kinds of mixed societies, including slavers and some F. sanguinea workers. They joined the crowd their slaves, are excellent for studying these phe­ of alien F. polyctena perfectly trustfully, without the nomena (nestmate-, kin- and species recognition, least hesitation or any signs of hostility and they acquisition and mixing of odours, stability of im­ tried to collect some pupae. However, the newcom­ printing and the odour acquired). It is just the ers recognized them immediately, attacked fiercely plasticity of callow ants that makes dulotic species - and killed all of them! successful (Le Moli 1980, Le Moli, Mori 1985). How­ ever, it is not only this. The extremely specialized slavers, ie amazon ants (eg Polyergus samurai DISCUSSION Yano), do not synthetize their own species-specific ^ CHC but they obtain them extemporaneously from The ability to distinguish between nestmates the slaves they have at a given time (Yamaoka and non-nestmates within the species and the 1990). ability to recognize foreign species are the basic Has F. sanguinea a possibility of such chemical element of the social organization in insects. Both mimicry? It seems that even it so, this possibility is depend, first of all, on colony-specific chemical sig­ limited. This is shown by the fate of the (ex-)slavers nals (and species-specific on the species level). which, after a year among a swarm of F. polyctena There still is a lot of ambiguity and controversy (SP-VI), were easily recognized and destroyed by

http://rcin.org.pl 24 • W. Czechowski

some other F. polyctena. It must be pointed out here their adult life, they recognized F. sanguinea brood that theF. sanguined individuals that had survived as foreign. until then were, most probably, the remainder of a This hypothesis has weak points. First, every­ new generation which had not been eaten complete­ thing took place inside natural nests of F. sangui­ ly by the slaves. Therefore they emerged among the nea, saturated with their specific odour, and under swarm of F. polyctena and as callow ants they were conditions of great overcrowding. Second, the slaves (theoretically) particularly liable to acquiring a manifested no aggressiveness towards adult work­ foreign odour. However, their own species-specific ers of the slavers which, too, frequently occurred in label appeared to be impossible to hide even in such proportionally insignificant numbers (eg SP-I, -II, an extreme situation. (Though it would be interest­ -VI, -IX, -X). And, after all, it is a rule in social ing to know how ants of typical slave species would Hymenoptera that alien brood is accepted more have behaved if they had been there instead of F. readily than adults (Holldobler 1977, Hólldobler, polyctena). Bergstrom and Lofquist (1968) have Michener 1980, Carlin 1988). (The last argument, found evidence of odour similarity between F. san- however, may be undermined by drawing attention guinea and some Seruiformica species but this to the activity of F. sanguinea workers in their general similarity may be a phylogenetic achieve­ contacts with young slaves - they help them during ment of the social parasite and thus an innate char­ emergence, feed them etc., and this means that their acter of the species. Moreover, these slave-keepers, mutual contacts are intensive). Third, the slaver quite unlike amazon ants, frequently live without brood was eliminated even in those cases when the slaves. And it is difficult to imagine a colony func­ percentage of slaves was not so high. Fourth, it did tioning without any discriminator. happen that, already after emancipation and in Slaves treat the nest of their slavers as their own. nearly monospecific colonies of F. polyctena, the This statement is a truism if it refers to the ants remaining F. sanguinea workers managed to have Seruiformica whose biology generally predestines their own offspring (males) which were looked after them to play the role of victims of social parasitism by F. polyctena (SP-VI). Fifth, what about slaver (eg dependent founding of nests by Formica s. str. queens? They, too, somehow fell a victim to the or Coptoformica Miill. queens). The fact that the slaves. mechanisms of chemical “indoctrination” work also Yet another explanation, though not fully satis­ on wood ants emphasizes the power of their impact. factory either, may be suggested. Now, as soon as Normally, wood ants and F. sanguinea seem to be the abundance af the colonies increased rapidly the the most fiercely fighting opponents in the entire ants were short of food. Hunger forced them to eat Palaearctic myrmecofauna. The same is empha­ their brood; this occurs frequently in ants (Wilson sized also by Marikovsky (1963). 1971). In an earlier paper (Czechowski 1990a) an The phenomenon of slave emancipation de­ assumption was put forward that each of the species scribed here raises some doubts in matters that - members of a given mixed colony selectively had have already seemed explained. First the question devoured the brood of the other and this had putF. arises: why the slaves, while approving of adult F. sanguinea, less abundant by far, in a hopeless posi­ sanguinea workers, destroyed the brood of the lat­ tion. However, it has been discoverd now that the ter. According to Jaisson’s suggestion (pers. comm.) offspring of the slavers disappeared even when the the possible cause lies in the quantitative dispropor­ slaves were in a minority. Therefore, there remains tion between the artificially introduced slave pupae an alternative explanation determined by the eco­ and the offspring of the slavers. It follows from this nomics of a mixed society as a whole: the younger interpretation that the slave callow ants (liable to brood were eaten. And, accordingly to the difference imprinting) that were emerging in great masses in the phenology of the two species, the brood of F. were mainly surrounded by a crowd of their con spe­ sanguinea were younger. cifics. During their early education they had not Even this hypothesis, however, does not explain much chance of getting in contact with a or a the causes and circumstances of the F. sanguinea pupa of the proper host of the nest. As a result, in queen elimination. It is possible that it was slaver

http://rcin.org.pl Emancipation of Slaves in Formica sanguinea Latr. Colonies • 25

queens who were intolerant (aggressive) of atypical not recognize alien F. polyctena as an enemy, but slaves and were killed by the latter in self- they themselves were perceived as such (SP-VI). defence(?). The camouflaging effect of another’s discriminator If, on the other hand, the reason for the elimina­ proved ineffective in this case. tion of the F. sanguinea brood was other than In other words: in a colony of F. sanguinea, F. hunger then the question is whether this phenome­ polyctena became, until its emancipation, «sangui- non was conditioned by the number (relative or nea» - in its own “feeling” and for the alien F. absolute) of the slaves. Was, thus, the elimination polyctena. After emancipation it again “became it­ of queens conditioned as well? And what is this self’ - for itself and for others. On the other hand, possible minimum number (or percentage) of slaves F. sanguinea after the emancipation of its slaves (from the point of view of F. sanguinea, the crucial became «polyctena» but only in its own “feeling”. permissible minimum)? The experiments have Why? Was this due to some kind of the influence of shown that both phenomena occur already when the the presence (or absence) of foreign (and their own) proportion of (atypical) slaves is about 50%. Does queens? this not mean that wood ants (as species) have at Some of the above questions may be explained if their disposal a peculiar safeguard aqainst social the experiments are repeated under laboratory con­ parasitism practised by F. sanguinea, and this ditions. Others require precise biochemical studies. makes the latter helpless against it? Let these speculations be finished with a reminder And another question still remains open: are of August Forel’s experiment, of last century, in typical slaves, ie the Serviformica species, capable which both the very idea and the result were amaz­ (even if only potentially) of do likewise? ing. (Forel repeated the experiment in many vari­ The last matter concerns the permanence of im­ ants with the same effect). He shovedF. sanguinea printing and an individual’s own and acquired odour and F. pratensis colonies into one bag, tied the bag discriminator. At the beginning of their life in a up, shook it vigorously and left for a few days. It mixed colony, highly dominated by slaves, these turned out that, after an initial massacre, the col­ very slaves - at the side of the relatively few slavers onies fused into one mixed society which remained - fought against their own conspecifics (even former in this form after the ants had been released (Forel nestmates) and the latter considered the former to 1874, 1928). be their enemies (SP-I, -II). The mechanism of this Acknowledgements. I thank Prof. Dr. J. Pętał, Prof. Dr. F. is well-known: on the one hand, imprinting of the Le Moli and Dr. J. Heinze for their comments on the manuscript odour of the slaver colony instead of their own and Mr. J. Murphy (British Arachnological Society, London) for colony- or species-specific label, and on the other, a checking and improving the English translation. camouflaging effect of this acquired odour against the inherited discriminator. However, after a year REFERENCES Bergstrom G., Lofquist J. 1968. Odour similarities between the (SP-VI), the same individuals (the new generation slave-keeping ants Formica sanguinea and Polyergus rufes- was to emerge later on) were friendly towards their cens and their slaves and . Journal of Physiology, 14: 995-1011. conspecifics and were treated likewise in return. Bonavita-Cougourdan A., Clćment J. L. 1986. Procćssus de rec­ This means that the original imprinting (from the onnaissance chez la fourmi Camponotus vagus Scop. Bulletin de la S. F. E. C. A., pp. 49-55. callow period) had died out and the influence of the Bonavita-Cougourdan A., Clćment J. L., Lange C. 1986. Nest- mate recognition: the role of cuticular hyarocarbons in the foreign odour was gone too (the latter is easy to ant Cam ponotus vagus Scop. Journal of Entomological Scien­ understand because of a change of nest and a mini- ces, 22: 1-10. Bonavita-Cougourdan A., Clćment J.-C., Lange C. 1988. Recon­ malized share of the former slavers). However, con­ naissance des larves chez la fourmi Cam ponotus vagus Scop. Phenotypes larvaires des spectres d’hydrocarbures cuticu- trary to what might have been expected the laires. Comptes Rendus des Sćances de 1’Acadćmie des Scien­ relations between F. polyctena and F. sanguinea ces, ser. III, 306: 299-305. Carlin N. F. 1988. Species, kin and other forms of recognition in after slave emancipation did not undergo a symmet­ the brood discrimination behaviour of ants. In: Advances in ric reversal. The last ex-slavers remainingfor a year myrmecology (ed. by J. C. Trager). E. J. Brill, Leiden, pp. 267-295. in the nest of the ex-slaves (and quantitatively domi­ Cool-Kwait E., Topoff H. 1984. Raid organization and behavioral development in the slave-making ant Polyergus lucidus nated by them even more than at the beginning) did Mayr. Insectes Sociaux, 31: 361-374.

http://rcin.org.pl 26 • W. Czechowski

Czechowski W., Pisarski B., Yamauchi K. 1995. Succession of ant Le Moli F. 1980. On the origin of slaves in dulotic ant societies. communities (Hymenoptera, Formicidae) in the moist conife­ Bolletino di Zoologia, 47: 207-212. rous forests. Fragmenta Faunistica (in press). Le Moli F., Mori A. 1982. Early learning and cocoon nursing Czechowski W. 1975a. Wyprawy rabunkowe Formica (R aptifor- behaviour in the red wood ant Formica lugubris Zett. (Hyme­ mica) sanguinea Latr. (Hymenoptera, Formicidae). Przegląd noptera: Formicidae). Bolletino Zoologia, 49: 93-97. Zoologiczny, 19: 33—43. Le Moli F., Mori A. 1984. The effect of early experience on the Czechowski W. 1975b. Mixed polycalic colony of Formica (Servi- development of “aggressive behaviour” in Formica lugubris formica) cinerea Mayr ana Polyergus rufescens Latr. (Hyme­ Zett. (Hymenoptera: Formicidae). Zeitschrift fiir Tierpsycho- noptera, Formicidae). Annales Zoologici, 33: 67-76. logie, 65: 241-249. Czechowski W. 1990a. Autonomization of slaves from mixed Le Moli F., Mori A. 1985. The influence of the early experience of colonies of Formica sanguinea Latr. and F. polyctena Foerst. worker ants on enslavement. Behaviour, 33: 1384- (Hymenoptera, Formicidae). MemorabiliaZoologica, 44: 55-63. 1387. Czechowski W. 1990b. A raid of Formica sanguinea L atr. (Hyme­ Le Moli F., Passetti M. 1977. The effect of early learning on noptera, Formicidae) on a conspecific colony. Memorabilia recognition, acceptance and care of cocoons in the ant Formi­ Zoologica, 44: 65-69. ca rufa L. Atti della Societa Italiana di Scienze Naturali e del Czechowski W. 1991a. Metoda “konia trojańskiego” - nowy spo­ Museo Civico di Storia Naturale, 118: 49-64. sób kolonizowania mrówek. Problemy, 6 (538): 34-36. Le Moli F., Passetti M. 1918. Olfactory learning phenomena and Czechowski W. 199lb. Mixed colonies of red wood ants (Hyme­ cocoon nursing behaviour in the antFormica rufa L. Bolletino noptera, Formicidae). Annales Zoologici, 44: 21—41. Zoolo'gica, 45: 389—397. Darwin C. R. 1859. On the origin of species by means of natural Marikovsky P. I. 1963. The ants Formica sanguinea Latr., as selection, or the preservation of favoured races in the struggle pillagers of Formica rufa Lin. nests. Insectes Sociaux, 10: for life. John Murray, London, 9 + 502 pp. 119-128. Dobrzański J. 1961. Sur 1* óthologie guerrióre de Formica san­ Móglich M., Hólldobler B. 1974. Social carrying behavior and guinea Latr. (Hymćnoptóre, Formicidae). Acta Biologiae Ex- division of labor during nest moving in ants. Psyche, 81: peri mentalis, 21: 53—73 . 219-236. Donisthorpe H. St. J. K. 1915. British ants, their life-history and Pisarski B., Czechowski W. 1990. The course of artificial coloni­ classification. W. Brendon and Son, Ltd., Plymounth, 15 + zation of red wood ants in the Gorce National Park. Mem­ 379. orabilia Zoologica, 44: 37^46. Forel A. 1874. Les fourmis de la Suisse. Socićtć Helvćtique des Rosengren R., Cherix D., Pamilo P. 1986. Insular ecology of the Sciences Naturalles, 4 + 452 pp. red wood ant Formica truncorum Fabr. II. Distribution, Forel A. 1928. The social world of the ants compared with that of reproductive strategy and competition. Mitteilungen der man. 2 vols. G. P. Putnam’s Sons, Ltd., London, 14 + 551, 20 Sen weizerischen Entomologischen Gessellschaft (Bulletin de + 445 pp. la Societe Entomologique Suisse), 59:63-94. Hólldobler B. 1971. Communication in social Hymenoptera. In: Stitz H. 1914. Die Ameisen (Formicidae). In: Die Insekten Mit- How Communicate (ed. by T. A. Sebeok). Indiana teleuropas insbesondere Deutschlands. II. Hymenopteren Univ. Press, Bloomington, pp. 418—411. (ed. bv Chr. Schroder). Stuttgart, pp. 1-111. Hólldobler B., Michener C. D. 1980. Mechanisms of identification Sudd J. H., Franks N., R. 1987. The behavioural ecology of ants. and discrimination in social Hymenoptera. In: Evolution of Blackie, Glasgow, London, 10 + 206 pp. Social Behavior: Hypotheses and Empirical Tests (ed. by H. Wheeler M. W. 1926. Ants, their structure, development and Markl). Verlag Chemie, Wainhaim, pp. 35-58. behaviour. Columbia Univ. Press, New York, 25 + 663 pp. Hólldobler B., Wilson E. 0. 1990. The ants. Belknap Press, Wilson E. 0. 1971: The insect societies. Belknap Press, Harvard Harvard Univ. Press, Cambridge, Mass., 12 + 732 pp. Univ. Press, Cambridge, Mass., 10 + 548 pp. Jaisson P. 1975. L’imprćgnation dans 1’ontogenćse des compor- Wiśniewski J. 1973. Badania nad przydatnością patentu PRL Nr temćnts aux cocons chez la jeune fourmi rousse ( Formica 54123 w sprawie pobierania poczwarek z mrowisk. Prace polyctena Fórst.). Behaviour, 52: 1-31. Komisji Nauk Rolniczych i Leśnych Poznańskiego Towarzy­ Jaisson P. 1980. Les colonics mixtes plurispćcifiques: un modele stwa Przyjaciół Nauk, 36: 241-250. pour l’ćtude des fourmis. Biologie et Ecologie Mćditer- Yamaoka R. 1990. Chemical approach to understanding interac­ raneenne, 7: 168—166. tions among organisms. Physiological Ecology, 27: 31-52. Jaisson P., Fresneau D. 1978. The sensitivity and responsiveness Yamaoka R. 1991. Importance of aggregation pneromone of ant of ants to their cocoons in relation to age and method of gueen for nestmate recognition: In: Evolution of Animal measurement. Animal Behaviour, 26: 1064-1071. Social Behaviour: JSPS-CNRS Seminar, Inuyama, pp. 34—36. Kutter H. 1968. Liste sozialparasitischer Ameisen. Archives Yamaoka R., Fujioka M., Yoshimura T., Hayoshiya K. 1988. Trimestrielles. Institut Grand-Ducal de Luxembourg, 33: Possibility of cuticular hydrocarbons as chemical cues of ants 201 - 210 . for recognition of nest mate, family and species. Proceedings 8-th International Congress of Entomology, p. 247.

http://rcin.org.pl